On the culture-specificity of linguistic gender differences: The case of English and Russian apologies1 EVA OGIERMANN Abstract The

present paper analyzes the influence of gender and culture on speech act performance. Although culture as a factor shaping gender roles can be regarded as being implicitly addressed by the growing number of speech act studies analyzing gender di.erences in various languages, results from such studies are dicult to compare. This study examines responses to o.ensive situations produced under identical contextual conditions by English and Russian women and men, thus ensuring comparability across groups and allowing for statements on the culture-specificity of linguistic gender di.erences. The first part of the paper o.ers a contrastive analysis of English and Russian responses to o.ensive situations, sketching some culture-specific di.erences in dealing with them. The second part is devoted to an investigation of gender-based di.erences in English and Russian, their comparison and interpretation. Apart from apology strategies and intensifying devices, the study also examines the use of downgrading strategies and the e.ect of strategy combinations on the illocutionary force of the responses. 1. Introduction 1.1. Feminist linguistics Ever since Lakos pioneering work on womens language, the field of feminist linguistics has been continuously gaining in popularity and broadening its scope. Early work described womens language as insecure and ineective (Lako 2004 [1975]) and linked linguistic gender dierences with unequal distribution of power (e.g., Thorne & Henley 1975; Fishman 1983). Whereas these studies have been criticized for portraying women as merely aberrant or defective copies of men (Bergvall 1999: 278), research into girls and boys socialization patterns no longer measured Intercultural Pragmatics 5-3 (2008), 259286 DOI 10.1515/IPRG.2008.013 1612-295X/08/00050259 6 Walter de Gruyter women against a male norm. It was shown that interactional styles are forged in early childhood and suggested that women and men are members of dierent sociolinguistic subcultures (Maltz & Borker 1982: 200). Numerous studies confirmed this theory by providing empirical evidence for gender-specific language usage (e.g., Tannen 1994; Holmes 1995; Coates 1994, 1996). It did not take long, however, before research focusing on dierences between female and male communicative styles met with criticism for ignoring diversity among women and similarities between women and men. In their seminal paper introducing the framework of communities of practice, Eckert and McConnell-Ginet (1998 [1992]) pointed out that generalizations about female and male language abstract gender from other aspects of social identity. Subsequent research increasingly challenged the concept of gender as a presupposed and stable category that determines linguistic choices. Instead, it came to be regarded as a dynamic concept realized through language use and interacting with other

aspects of social identity in particular communities of practice. In the attempt to move beyond the binary oppositions of male and female (Mills 2003: 169) and break down destructive dichotomies (Freed 2005: 704), the field has inevitably shifted its focus from the prototypical to the untypical and exceptional (see e.g., McElhinny (1998) on female police ocers or Bucholtz (1999) on female nerds). It is, therefore, not surprising that several researchers voice the need for studies establishing gender-specific patterns, while arguing that they are indispensable in identifying and interpreting individual uses (e.g., Bergvall 1999: 288, Holmes 2005: 111). Bergvall maintains that the study of language and gender involves the analysis of what is innate, what is socially constructed locally, and what is ideologically constructed (1999: 285). Dierentiating between these aspects of gender identity clearly illustrates the limitations of the communities of practice approach, and even linguists adopting this framework identify constraints on what linguistic and pragmatic resources are available to women as opposed to men (Christie 2005: 4). 1.2. Feminist linguistics in Russia? Western feminists who took interest in Eastern Europe were disappointed to see that most East European women reject the ideas of the feminist movement (Temkina & Rotkirch 1997). In a survey conducted in Poland, some women even pointed out that there were more important problems in need of reform in Poland than language (Koniuszaniec & Baszkowska 2003: 277). 260 Eva Ogiermann While women in the West were fighting for equal opportunities, those behind the Iron Curtain were emancipated to subordination (Temkina & Rotkirch 1997) by socialist ideology. As early as 1901, twenty-six percent of Russian industrial workers were women, while in 1917 the female proportion of the workforce reached forty percent (Comrie & Stone 1978: 160). Women in the Soviet Union were not only expected to be both workers and mothers, but also had to spend a considerable amount of their time queuing in front of badly supplied shops to be able to feed their families. The introduction of a market economy has not improved the situation of Russian women. It deprived working mothers of the help the state used to provide in child care, and the discrepancy between male and female salaries has considerably increased (Azgichina 2000). Russian linguists take a rather critical attitude towards Western work on language and gender, arguing that, having originated in the feminist movement, this discipline is necessarily subjective (Zemskaja et al. 1993: 94). Studies discussing the relationship between gender and the Russian language tend to focus on female forms of professional terms (e.g., Comrie & Stone 1978; Tafel 1997) and the representation of women and men in proverbs and idioms (Tafel 1997; Doleschal & Schmid 2003). The edited volume entitled Slavic Gender Linguistics (1999)a collection of papers examining linguistic gender dierences in several Slavic languagesremains an exception. In the introduction to the book, the field is described as one that has become a research focus for Russian and Western Slavic linguists (M. H. Mills 1999: vi), but little has been

published on the subject since. Russian studies examining gender-specific language use are largely limited to those conducted by Zemskaja and her colleagues (e.g.; Zemskaja et al. 1993). What makes Zemskajas work exceptional is that she started assembling a corpus of spoken Russian in the 1970sat a time when most Western researchers relied on introspection, and in a country where, up to the present day, linguists show little interest in spoken language. Although she does identify dierences between female and male speech in her data, she regards them as negligible and defines the preferences for certain grammatical forms by one of the genders (e.g., diminutives by females) as mere tendencies of language usage. She does not attribute them to gender per se but to the dierent roles women and men fill within Russian society: The dual role of a working mother, for instance, enables Russian women to split their attention and switch between topics more easily than men. Zemskaja maintains that factors such as the speakers age, character, education, profession and social role are more decisive in shaping their On the culture-specificity of linguistic gender di.erences 261 speech than gender, and she rejects the idea that Russian women and men have distinct conversational styles (1993: 132). 2. The present study Although gender has been recognized to interact with other aspects of social identity, such as class, age, and culture, most feminist linguists analyze the language of Western, middle class women (but see Bucholtz 1996; Mendoza-Denton 2004; Goodwin 2005). The present study compares gender-specific language use in two countries: Britain, where, partly as a result of the feminist movement, gender roles have considerably changed over the past decades, and Russia, where linguistic gender dierences have never been perceived as significant or problematic. Accordingly, the study illustrates the importance of culture as a factor responsible not only for dierences between female and male conversational styles but also for attitudes towards these dierences. Whereas significant gender-specific dierences in both languages would support Western research, their lack would confirm Zemskajas doubts concerning the objectivity of Western feminist linguistics. A marked discrepancy in the extent to which gender-specific dierences materialize in English and Russian, in contrast, can be attributed to cultural factors. 2.1. Gender and apologies The present study focuses on gender- and culture-specific use of apologies. This speech act, perhaps due to its vital social function of restoring and maintaining social harmony, has received ample attention in previous research. It was one of the speech acts examined in the Cross-Cultural Speech Act Realization Project (Blum-Kulka et al. 1989), and the framework developed in this project was adopted in numerous studies analyzing apologies in various languages. Despite the recent criticism of Brown and Levinsons (1987) speech act theoretical approach to politeness (e.g., Eelen 2001; Mills 2003; Watts 2003), apologies remain a popular and insightful unit of analysis, as evidenced by an issue of the Journal of Politeness Research (2007) devoted entirely to this speech act. Previous studies on gender-specific apology behavior, analyzing various

languages and types of data, tend to agree in reporting dierences between female and male apologies, while generally portraying women as more apologetic than men. The most influential contribution to the study of gender-specific use of apologies has been made by Holmes (e.g., 1995). In her corpus of ethnographically collected apologies oered by New 262 Eva Ogiermann Zealand women and men, seventy-five percent of all apologies were offered by women and seventy-three percent to women. Holmes suggests that apologies have gender-specific functions: while for women, they express solidarity and concern for others, men regard them as admissions of weakness, inadequacy or failure (1995: 175). Tannens analysis of conversations at work (1994) also shows American women more willing to apologize than men, and so does Meyerhos study of a speech community of Vanuatu (1999). In her data, the word sore (sorry) occurs almost exclusively in female speech, where it assumes a wider range of functions than it does in male speech. The multifunctional nature of apologies has also been illustrated by Christies analysis of parliamentary debate (2005). Although British male Members of Parliament (MPs) apologized more often than female MPs, they mainly did so in order to be able to perform a Face-Threatening Act (FTA). The function of the few apologies uttered by female MPs, in contrast, was to take responsibility and express concern (2005: 24). Similarly, Bean & Johnstone (1994) found that American males apologized twice as often as females during telephone interviews, where apologies served the function of discourse managing devices. Researchers analyzing apologies on the basis of linguistic corpora (e.g., Aijmer 1996; Deutschmann 2003) do not report significant genderdi erences in the use of apology formulae. It should be borne in mind, however, that while formulaic apologies can be easily located in a corpus, it is virtually impossible to identify indirect apologies and oenses for which no apology was oered, making it dicult to compare apology behavior across groups (compare Ogiermann 2004). Incidentally, this problem also aects the reliability of observational data (Holmes 1995: 157), but it can be avoided by employing experimental data elicitation techniques, such as role-plays and written discourse completion tests (DCTs). Studies based on role-play data, e.g., Cordellas study of apologies in Chilean Spanish and Australian English (1990) and Marquez Reiters analysis of British English and Uruguayan Spanish (2000) report an overall higher frequency of apology strategies in the female data. DCT studies confirm this tendency, and since they generally rely on large samples of data, the findings allow for fairly reliable generalizations about the apology behavior of the examined groups. Studies using DCT data show that women employ more apology strategies than men in a variety of languages, among them: American English (Bataineh & Bataineh 2005), British English (Ogiermann 2002), Peninsular Spanish (Stapleton 2004), Russian (Shardakova 2005), Hungarian (Suszczynska 2005), and Jordanian Arabic (Bataineh & Bataineh 2006). On the culture-specificity of linguistic gender di.erences 263 3. Method

Although the studies discussed above agree in portraying women as more apologetic than men, the results are dicult to compare, since the reported dierences are not always equally significant. Moreover, these studies look at dierent communities of practice, dierent functions of apology formulae, and use dierent data collection methods. The large corpora of comparable and replicable data that can be assembled by means of a DCT make this method indispensable for contrastive pragmatic research. What makes this instrument particularly useful for the present study is that data produced under identical contextual conditions by women and men in dierent languages can yield valuable insights into the culture-specificity of linguistic gender dierences. 3.1. Data collection instrument Despite the criticism DCT data have received for not adequately reflecting natural speech, there seems to be a consensus that DCT responses accurately reflect the content expressed in natural speech and the values of the native culture (Beebe & Cummings 1996: 75) and indicate what strategic and linguistic options are consonant with pragmatic norms (Kasper 2000: 329). Kasper even suggests that authentic data may just not be a viable option when (. . .) the research goal is to compare the use of specific pragmatic features by dierent groups of speakers (320). The DCT developed for this study consists of ten scenarios, eight of which depict oensive situations, while two serve as distracters. The scenarios include selected combinations of the variables of social distance and power and hearer gender2, so that the situations describe encounters with friends, strangers, socially more powerful acquaintances, and authority figures of both genders (see appendix for a copy of the questionnaire). 3.2. Subjects The subjects participating in the study were all studentsa group sharing various practices and fairly homogenous in terms of age, occupation, education, and social class. Since this description applies to British as well as Russian students, the present study can be said to compare similar communities of practice in dierent cultures. Although students constitute a mixed-gender community of practice, it has been long acknowledged that academic mens language is not 264 Eva Ogiermann representative of what is generally associated with typical male speech (Lako 2004: 47 [1975]). In addition, the data were collected in departments with high proportions of female students, where gender dierences can be expected to be even less significant than in other academic communities of practice. The English DCT was distributed at universities in London, Cardi, and Swansea, and the Russian version at two universities in Moscow. The corpus is comprised of 100 English and 100 Russian DCTs, with equal distribution between genders. Accordingly, the data consist of a total of 1600 responses elicited under identical contextual conditions from comparable population groups. The average age is 20.4 for the British and 17.9 for the Russian subjectsthe younger age of the Russian students being due to the fact that they can enter university at the age of 16.

4. A contrastive analysis of English and Russian apologies An analysis comparing gender-dierences in English and Russian apology realizations cannot be conducted without knowing what it means to apologize in these languages. Hence, the present chapter is devoted to briefly outlining the dierences between English and Russian apologies established in the data. Simultaneously, it describes the taxonomy used to categorize the data, which is a modified version of Olshtain and Cohens speech act set of apologizing (1983). 4.1. Direct apology strategies The most evident dierences between Russian and English apologies appear in connection with linguistic realizations of Illocutionary Force Indicating Devices (IFIDs). Although both languages oer several formulaic expressions potentially serving as apology strategies (for a detailed discussion see Ogiermann 2006, 2007), each language exhibits a marked preference for one IFID realization. My English data display an extraordinarily strong focus on the expression of regret, with 635 out of 645 IFIDs ( 98%) taking the form of (Im) sorry. The full form Im sorry is slightly more frequent, amounting to 336 instances, while the short form sorry occurs 299 times. The most conventionalized Russian IFID is the request for forgiveness, which is performed by means of the verbs izvinit and prostit (to forgive). Izvinit contains the word vina (guilt), so that by using the imperative izvini (T-form) or izvinite (V-form), the speaker literarily asks to be freed from guilt. The formula prosti-teetymologically going back to the On the culture-specificity of linguistic gender di.erences 265 permission to stand up after bowingis associated with religious contexts and forgiving sins. This might be the reason why it is used to apologize for more serious, even unpardonable oenses (Rathmayr 1996: 66). In my data, izvini-te is more frequent than prosti-te, with 404 occurrences of the former and 125 of the latter. Among the remaining IFID realizations, there are 27 apology expressions containing infinitive forms of these two verbs or nouns derived from them, so that they can be regarded as stylistically marked variants of the request for forgiveness. With a total of 586 IFIDs in the Russian data, the various forms of the request for forgiveness make up ninety-five percent of all IFIDs. Although IFIDs are highly routinized and speakers employing them are more likely to be aware of their function than their meaning, the focus on one IFID form in a particular language allows an insight into its culture-specific concept of apologizing. In English, apologies are generally viewed as formulae expressing regret, and have thus been assigned to the category of expressives. In Russian, the concept of guilt is central to the speech act of apologizing, and since the main IFID realization has the form of a request, Russian apologies have been classified as belonging to the category of directives (Rathmayr 1996). Another factor distinguishing requests for forgiveness from expressions of regret is the degree to which they threaten both interlocutors face. In requesting forgiveness, the oender not only places her- or himself at the hearers mercy, but also threatens the hearers negative face by assigning her or him an active role in the process of forgiving. The expression of regret, on the other hand, implies a much lower degree of face-threat for

both parties involved. The focus on this IFID in English seems to be in accordance with the preference for negative politeness in Anglo-Saxon culture (Brown & Levinson 1987), whereas the use of requests for forgiveness in Russian confirms the classification of Russian culture as a positive politeness culture (Rathmayr 1996). 4.2. Intensification Since the expression of regret is relatively weak in its apologetic force (Suszczynska 1999: 1060), it requires more intensification than other IFIDs. Another reason why the Russian subjects used fewer intensifiers than did the British ones might be that the imposition inherent in the request for forgiveness increases with intensification. The total number of intensifiers in the English data amounts to 273 instances, which means that the English respondents intensified fortytwo percent of all their IFIDs. The frequency of intensifying devices in Russian is considerably lower, for they come to only 136 instances, thus 266 Eva Ogiermann accompanying twenty-three percent of Russian IFIDs. A chi-square analysis shows this dierence to be highly significant, but one should bear in mind that the intensifying devices preferred in the two languages are not fully comparable. While nearly all English intensifiers take the form of adverbials, such as really, so and very, in the Russian data, due to the strong focus on the request for forgiveness, eighty-six percent of intensifying devices are represented by the word pozalujsta (please). Exclamations also have an intensifying function, though they primarily serve as expressions of surprise. They are more frequent in the English data, where they occur 166 times and center around expressions such as oh, oh shit, oh my God, oops, etc. The Russian subjects used 112 exclamations, most of them taking the form of oj, occasionally oops and gospodi (Lord). However, since the data have been elicited by means of a written DCT, which is not a reliable instrument for examining linguistic devices expressing spontaneity and emotionality, these figures should be regarded as indicative only. 4.3. Indirect strategies Whereas strong agreement on a particular IFID realization among speakers of one language illustrates the culture-specificity of formulaic apologies, indirect strategies are more likely to reflect conscious pragmatic choices. Although the classification of the data largely adheres to the speech act set of apologizing suggested by Olshtain and Cohen (1983), some changes were considered necessary to meet the requirements of this study, one of them being the addition of the strategy concern for hearer. A more substantial change consists in replacing the strategies explanation and taking on responsibility by a group of account strategies categorized according to the degree of responsibility acceptance and the corresponding face-threat inherent in them. 4.3.1. Positive politeness apology strategies. The strategies oer of repair, promise of forbearance and concern for hearer are all contextually restricted to particular oensive situations. Repair is generally Table 1. Total numbers of IFIDs and intensifying devices in English and Russian N 800 English Russian w2 (df 1)

IFIDs 645 586 2.74 n. s. Intensifiers 273 136 45.22 p < .0001 Exclamations 166 112 10.01 p < .0015 On the culture-specificity of linguistic gender di.erences 267 oered in situations in which material damage has been caused, forbearance is an important element in apologies for recurrent oensive behavior, and concern for hearer is likely to be expressed when physical or psychological damage may have occurred. What these three strategies have in common is not only that they are indirect realizations of the speech act of apologizing, but they also exhibit a stronger orientation towards positive face needs than IFIDs. Incidentally, they all appear in Brown and Levinsons chart of positive politeness strategies as oer, promise and attend to H (1987: 102). Considering that Russian has been classified as a positive politeness language while British English favors negative politeness, the preferences for the three positive politeness strategies are surprisingly similar, concern for hearer even being significantly more frequent in the English data. Most oers of repair consist of formulations in the future tense, naming the measures that will be taken in order to compensate for the damage, e.g., Ja kuplju tebe novych rybok (I will buy you new fish). These direct realizations express a high degree of obligation on the part of the speaker. Indirect realizations of oers of repair, in contrast, generally take the form of questions and suggestions, such as Would you like me to tidy up? or Mozno ja kuplju tebe drugich? (May I buy other ones for you?). Indirect oers can convey consideration for the hearers wishes but also reluctance, depending on the exact formulation or prosodic features, but also the interlocutors cultural background. Although one would expect a stronger preference for direct forms in a positive politeness and indirect forms in a negative politeness culture, direct oers of repair are only slightly more frequent in the Russian data, where they amount to 289 instances as opposed to 274 in the English data. Indirect realizations, in contrast, were used by 91 English and 92 Russian respondents. However, dierences do appear in connection with the use of intensifying devices emphasizing the willingness to repair the damage. Whereas 40 oers of repair formulated by Russians include the adverb objazatelno (definitely), in the English data, there is only one instance of definitely and two of of course. Table 2. Total numbers of positive politeness apology strategies in English and Russian N 800 English Russian w2 (df 1) Oer of repair 365 381 0.3 n. s. Promise of forbearance 47 53 0.26 n. s. Concern for hearer 55 20 15.42 p < .0001 268 Eva Ogiermann Furthermore, both sets of data include formulae aiming at appeasing the hearer, which can also be defined as positive politeness strategies. In the English data, there are 16 instances of the expression dont worry and one of no worries. The Russian respondents, in contrast, used 30

appeasing formulae consisting of negated imperatives of four synonymous verbs: expressing the state of being restless (bespokoitsja and volnovatsja), the process of worrying ( perezivat) and getting in a bad mood (rasstraivatsja). Hence, Russian not only shows a stronger preference for this positive politeness strategy, but also makes use of a broader range of formulae. Generally, there seems to be a tendency for the British to use routinized formulae and for Russians to favor individually phrased strategies. When promising forbearance, 33 English and 33 Russian subjects used stereotypical phrases applicable in any oensive situation with a potentially recurring character, such as: Bolse etogo ne povtoritsja/It wont happen again. Promises of forbearance explicitly addressing the circumstances of the oense, in contrast, appear 20 times in the Russian and 14 times in the English data. The English speakers preference for formulaic expressions is even more marked in the case of concern for the hearer. Although the English informants used nearly three times as many expressions of concern as did the Russians, most of them consisted of two variants of only one formulaic expression, namely Are you OK? and Are you alright?. The Russian respondents, in contrast, resorted to various individual formulations aimed at ascertaining that the victim was OK (S vami vse v porjadke?) didnt get seriously hurt (Vy ne silno udarilis?), and asking how she felt (Kak vy sebja cuvstvuete?). Consequently, although the Russian data exhibit a relatively low frequency of expressions of concern, the linguistic realizations are more varied than the English ones, showing Russians more attentive to contextual features. 4.3.2. Accounts. The category of accounts encompasses the strategies explanation and taking on responsibility originally distinguished by Olshtain and Cohen (1983). The reason why this distinction was abandoned is that it has proved problematic in previous research, for these strategies tend to overlap: While expressions of responsibility often refer to the circumstances of the oense, explanations tend to reveal the speakers willingness to accept responsibility for the situation. Furthermore, discussions of taking on responsibility often include strategies downgrading and denying responsibility, thus covering the entire spectrum of strategies related to responsibility. What is generally overlooked, though, is that explanations may also refer to circumstances On the culture-specificity of linguistic gender di.erences 269 either linking the speaker with the oense or delegating the responsibility elsewhere. Therefore, these two strategies have been put together and divided into upgrading and downgrading accounts. Table 3 provides examples illustrating the account types distinguished in the present study. Although they have been placed on a continuum of increasing responsibility acceptance and face-threat, the exact order is Table 3. Account types English examples Russian examples Downgrading Deny It wasnt me.S.6. I did return them on time.S.2 Ja tut ni pri cem.S.1

[lit. I am here not with anything.] Obratites v sosednjuju dver.S.6 [Turn to the neighboring door.] Act innocently Ive no idea what happened.S.1 I just found them floating in the morning.S.1 Ponjatija ne imeju!S.1 [I have no idea!] Ne znaju kak e to ko mne popalo. S.7 [I dont know how this got to me.] Minimize It was only one party. It doesnt happen that often.S.6 Raz v godmozno.S.6 [Once a yearits allowed.] Excuse The ticket machine had broken in the station.S.6 C to-to sluciloss zamkom.S.1 [Something has happened to the lock.] Admit facts Your fish died.S.1 Tvoi rybki sdochli.S.1 [Your fish died.] Upgrading Justify I didnt realize how frail they were.S.1 Ja prosto vas ne zametila.S.4 [I simply did not notice you.] Lack of intent It was a genuine mistake.S.7 I had no intention of stealing it.S.7 E to vyslo slucajno.S.4 [This happened accidentally.] Embarrassment This is really embarrassing.S.5 Mne tak neudobno.S.2 [lit. This is so unpleasant to me.] Accept That was totally my fault.S.2 I dont think I fed them enough.S.1 Ja vinovata pered toboj.S.2 [I am guilty before you.] Ja plocho uchazival za rybkami. S.1 [I badly looked after the fish.] Self-Criticism Im completely useless.S.2 Ja byla ocen neostorozna.S.4 [I was very inattentive.] 270 Eva Ogiermann open to re-arrangement, being dependent on the exact linguistic realization and contextual factors. While excuses, admissions of facts, and justifications necessarily refer

to the circumstances of the oense, the remaining account strategies comprise formulaic expressions focusing on the oenders responsibility as well as situation-specific realizations focusing on the contents of the oense. Downgrading accounts include strategies denying responsibility directly, e.g., by negating ones involvement in the oense, and indirectly, e.g., by acting innocently. The category further contains expressions minimizing the oensiveness of the situation and excuses providing external factors leading up to the oense. Admissions of facts can be defined as distancing devices referring to the oense in a neutral way. By resorting to this strategy, the speaker neither accepts responsibility nor attempts to reduce or deny it. Although justifications name circumstances which are face-saving for the speaker, they do not deny responsibility for the oense and generally add to the illocutionary force of the apology. Many of the realizations of this account type emphasize the accidental nature of the oense, which is exactly the function of the slightly more face-threatening account type termed lack of intent. What distinguishes this strategy from justifications is that it does not contain any information making the oense more pardonable. Expressions of embarrassment indirectly convey a negative evaluation of ones behavior, whereas direct admissions of responsibility clearly link the speaker with the oense. Responsibility can be accepted by means of formulaic expressions of guilt as well as confessions critically portraying the oenders behavior. Expressions of self-criticism only indirectly accept responsibility, but they are highly face-threatening for the speaker and clearly have an upgrading function. The frequency of upgrading accounts in the two languages is surprisingly similar, with 605 instances in the English and 607 in the Russian data. Furthermore, justifications, the least face-threatening and most frequent upgrading accounts, occur 440 times in the English and 441 times in the Russian data. The only notable dierence is that the Russian respondents show a stronger preference for expressions of lack of intent and the English subjects for expressions of embarrassment (see table 7 for all frequencies). Downgrading accounts, in contrast, are significantly more frequent in the English data, where they amount to 294, as opposed to only 182 instances in the Russian data (w2 25:88 [df 1], p < .0001). This significant disparity is mainly due to a higher frequency of accounts On the culture-specificity of linguistic gender di.erences 271 admitting facts, acting innocently and denying responsibility in the English data. Consequently, the main dierence between the two languages in the use of account strategies is that Russians seem less protective towards their face when apologizing. This is not only reflected in the lower frequency of downgrading accounts in the Russian data but also in the choice of linguistic realizations. Whereas British responses tend to consist of brief expressions, such as: I forgot, Russian responses are more likely to contain an explicit reference to the oense, as in: Ja sovsem zabyla, cto disk ostalsja u menja v rukach (I totally forgot that the CD has remained in my

hands). 4.4. Strategy combinations The inclusion of downgrading accounts in the analysiseven though they do not serve the function of apology strategiesproved indispensable for the interpretation of the data. Since the English data contain more IFIDs and more downgrading accounts than do the Russian responses, one could hypothesize that combinations of these two contradictory strategies are more likely to occur in English. Moreover, the expression of regret can easily merge into a mere expression of sympathy, as in: Im sorry, Ive no idea what happened. Hence, comparisons of total numbers of particular strategies across groups are problematic in that they do not account for the eect strategy combinations may have on the illocutionary force of the entire response. Since responsibility acceptance is an indispensable element of an apology (see, e.g., Fillmore 1971: 287 or Holmes 1990: 161), the responses were classified into four categories according to two factors: whether they contain an IFID and whether they accept or deny responsibility (see table 4). Whereas an apology formula accompanied by strategies denying responsibility will not result in a successful apology, combinations of upgrading accounts with one of the positive politeness strategies generally do, even though they do not include an IFID explicitly marking the response as an apology.3 The distribution illustrated in table 4 exhibits a significant dierence between the English and the Russian data (w2 33:62 [df 3], p < .0001). Table 4. IFIDs vs. responsibility acceptance in English and Russian I FID/ esp _IFID/ esp FID/_Resp _IFID/_Resp Total R R I English 507 81 81 131 800 Russian 506 150 56 88 800 272 Eva Ogiermann The most remarkable result of this classification is that although IFIDs are more frequent in the English data, the frequencies of apologies including an IFID ( FID/ esp) are almost identical in both languages, I R and the total number of responses accepting responsibility ( FID/ I R esp & _IFID/ esp) amounts to 656 in Russian and to only 588 in R English. In other words, whereas my Russian subjects show a greater tendency to apologize without resorting to an IFID, the British produce more responses in which the apologetic attitude conveyed by an IFID is cancelled out by the addition of face-saving strategies. 5. A contrastive analysis of male and female apologies in English Before analyzing the apology strategies employed by English women and men, I would like to devote some attention to a stereotype associated with female speech, namely that women talk more than men. Provided that elaborate apologies are more polite than short ones, the length of the responses may indicate that women are particularly concerned about maintaining social harmony. My data seem to confirm this hypothesis as the total amount of words is 5931 in the female and only 5067 in the male data. However, one should be careful when interpreting this significant dierence (w2 67:72 [df 1], p < .0001) since the data also include downgrading strategies.

The distribution of IFIDs confirms the tendency for women to formulate more explicit apologies than men, for the number of IFIDs amounts to 348 instances in the female and to only 297 instances in the male data (w2 3:88 [df 1], p 0.0489). An additional distinction that can be made is that between short and long forms of the expression of regret, the elliptic character of the former making them slightly less apologetic. The male subjects show a stronger preference for short forms: There is a total of 164 instances of the elliptic sorry and only 128 of the full form in the male data. The female respondents, in contrast, used 208 full forms and 138 short forms. This significant preference (w2 13:76 [df 1], p 0.0002) for long forms in the female data becomes even more important when considering that this variant of the expression of regret is more likely to be intensified. Accordingly, the preference for full forms in the female data is paralleled by a high frequency of adverbial intensifiers: Whereas women employed 184 intensifiers, men used only 89 (w2 32:36 [df 1], p < .0001). The dierences in the repertoire of intensifying devices, in contrast, are rather negligible. Whereas the range of adverbial intensifiers in the male On the culture-specificity of linguistic gender di.erences 273 data was restricted to the four variants: really, so, very and terribly, female respondents occasionally also used truly and ever so. The frequencies of exclamations are even more discrepant. With 118 instances in the female and 48 in the male data (w2 28:68 (df 1), p < .0001), British women not only seem to be more concerned about their apologies sounding particularly sincere, but also appear more emotional than British men. The three positive politeness strategies are all slightly more frequent in the female data, but none of the dierences reaches statistical significance (see table 7). An analysis of oers of repair according to the level of directness underlying their linguistic realizations indicates that, contrary to common expectations, British women tend to use more direct forms than men. Whereas direct realizations occur 147 times in the female and 127 times in the male data, indirect formulations were favored by 46 British women and 45 British men. A possible reason for the slightly higher degree of directness in the female data is that oers are beneficial to the hearer. The gender-specific preferences for appeasers, however, do not render the female respondents particularly hearer-supportive, for these linguistic devices are nearly equally frequent in both sets of data. A look at the distribution of upgrading and downgrading accounts in the English data shows women a little more willing to accept responsibility and risk losing face than men. Upgrading accounts occur 285 times in the male and 320 times in the female data, whereas downgrading accounts amount to 155 and 139 instances, respectively. Since the slightly higher number of upgrading accounts in the female data is mainly due to a stronger preference for justificationsthat is for the least facethreatening upgrading account typethis dierence can be dismissed as negligible. The higher frequency of expressions of embarrassment in the female data, in contrast, can be interpreted as confirming previous findings portraying female apologies as more other-oriented than mens (e.g., Holmes 1995).

Further interestingeven though not statistically significantdifferences materialize in the use of downgrading accounts, one of them being that men seem to be more inclined to provide excuses than women. When refusing to accept responsibility, men favored direct denials while women resorted more often to indirect realizations of this strategy. Their preference for a more defensive way of denying responsibility may be interpreted as an example of female lack of assertiveness. An analysis of combinations of IFIDs and strategies accepting and denying responsibility shows that apologies including an IFID are slightly more frequent in the female data, while the numbers of responses that cannot be defined as apologies are similar for both groups. 274 Eva Ogiermann Although the dierent distributions in the female and male data do not reach statistical significance, an interesting pattern emerges in connection with responses that are likely to be accepted as apologies even though they do not include an IFID and those denying responsibility despite the inclusion of an IFID. British men apologized without resorting to routinized apology formulae more often than women, while the frequencies of responses combining an IFID with an unapologetic attitude are comparable in both sets of data. As a result, although IFIDs were used significantly more frequently by British women, the numbers of responses classifying as apologies are similar for both genders, as they amount to 297 in the female and 292 in the male data. 6. A contrastive analysis of male and female apologies in Russian Not surprisingly, the stereotype of women being more talkative than men also exists in Russia (Zemskaja et al. 1993), and the data indicate that Russian and British women are very similar in this respect. Incidentally, the discrepancy between the lengths of responses formulated by male and female Russians is nearly identical with that established for the English data. In both languages, fifty-four percent of the words were uttered by females and forty-six percent by males. Since the Russian responses are generally shorter, the total number for each gender is lower, with the female data comprising 4,094 and the male 3,536 words (w2 40:66 [df 1], p < .0001). The number of IFIDs is slightly higher in the female data, where it amounts to 308 instances. With a total of 278 IFIDs used by Russian males, this discrepancy does not reach statistical significance. As I have argued earlier, requests for forgiveness are the preferred Russian IFID realization, making up ninety-five percent of all IFIDs in the data. The imperative form of the verb izvinit occurs 206 times in the female and 198 in the male data. The overall higher frequency of IFIDs in the female data is, therefore, mainly due to preferences for the form prosti-te, which was used by 80 females and 45 males (w2 9:24 [df 1], p 0.0024). Table 5. IFIDs vs. responsibility acceptance in the English data I FID/ esp _IFID/ esp FID/_Resp _IFID/_Resp Total R R I Female 266 31 41 62 400 Male 242 50 39 69 400 On the culture-specificity of linguistic gender di.erences 275 Analogously to the divergent illocutionary forces of the full and

short forms of the English expression of regret, the two verbs available for performing the request for forgiveness in Russian have dierent implications. The religious connotations of the verb prostit make it more self-humiliating than the more common izvinit. Hence, both British and Russian women favor the apology form with the stronger illocutionary force. However, out of the 27 instances of the stylistically marked requests for forgiveness found in the Russian data, 20 were employed by men. Although Russian women used more intensifiers than Russian men, the discrepancy is less marked than in the English data. Out of the 136 intensifiers found in the Russian material 87 were used by women and 49 by men (w2 10:06 [df 1], p 0.0015). Another element occasionally accompanying the request for forgiveness with a potentially intensifying function is the direct object menja. Whereas English requests for forgiveness normally include the pronoun me, identifying the speaker as the person to be forgiven, this element is often omitted in Russian. According to Rathmayr, the added menja makes the expression sound more serious (das hinzugesetzte menja verleiht der Auserung Ernst) (1996: 72), whereas Formanovskaja argues that it makes the formula (. . .) more personal (delaet formulu (. . .) bolee licnostnoj) (2002: 125). In either case, it can be regarded as a form of intensification. In the present data, menja occurs 38 times, that is with only seven percent of all requests for forgiveness uttered by my Russian respondents. Interestingly, this intensifying device was used almost exclusively by women, with only six instances in the male data. Rathmayrs claim that the exclamation oj is characteristic of female speech (1996: 92) cannot be confirmed by my results, though the dierence between the 72 exclamations appearing in the female and the 40 instances in the male data is statistically significant (w2 8:58 [df 1], p 0.0034). The distribution of positive politeness apology strategies in the Russian data mirrors the pattern emerging from the English data, with females exhibiting a slightly stronger preference for positive politeness strategies than males. The most marked dierence arises in connection with expressions of concern, whose low occurrence, however, does not allow for generalizations (see table 7). Dierences concerning linguistic realizations of the three positive politeness strategies are greatest in the case of oers of repair, the distribution of direct and indirect formulations resembling that in the English data. Whereas Russian men used 139 direct and 43 indirect realizations, in the female data these figures amount to 150 and 49, respectively 276 Eva Ogiermann again showing women slightly more direct when performing a speech act which is beneficial to the hearer. Disparities in the use of intensifiers and appeasing formulae accompanying oers of repair further confirm this tendency. The intensifying adverbial objazatelno appears 26 in the female and only 14 times in the male data, showing Russian women more determined to compensate for the damage. Appeasers were also more popular among Russian women, for they employed them 18 times, while resorting to all four verbs identified

in the data. In the male data, in contrast, there are only twelve instances of three variants of appeasers. The distribution of account strategies resembles that established for the English data. Russian men used 281 upgrading and 99 downgrading accounts whereas women used 326 of the former and 83 of the latter, thus showing an overall greater willingness to accept responsibility. This is mainly due to the statistically significant dierences in the distribution of excuses and expressions of lack of intent, while the frequencies of the remaining account strategies are comparable. Excuses appear 45 times in the male and 25 in the female data (w2 5:16 [df 1], p 0.0231), fully confirming the pattern emerging from the English data. Expressions of lack of intent, in contrast, were used 59 times by women and 26 time by men (w2 12:04 [df 1], p 0.0005), which stands in opposition to the distribution in the English data, where men expressed lack of intent more often than women. A classification of the Russian responses according to whether they contain an IFID and whether they accept responsibility yields findings partly confirming those emerging from the English data. Russian responses including an IFID are slightly more frequent in the female data, where they amount to 294 instances, as opposed to 268 in the male data. The total numbers of responses accepting responsibility are even closer, with 331 instances in the female and 325 in the male data, thus considerably narrowing down the discrepancy in the use of IFIDs. On the whole, Russian men and women show a stronger agreement in apologizing without resorting to a formulaic apology strategy than do the Table 6. IFIDs vs. responsibility acceptance in the Russian data I FID/ esp _IFID/ esp FID/_Resp _IFID/_Resp Total R R I Female 261 70 33 36 400 Male 245 80 23 52 400 On the culture-specificity of linguistic gender di.erences 277 English respondentsthough Russian men tend to deny responsibility in an unambiguous way, whereas women seem more inclined to disguise their unapologetic attitude by using an IFID. 7. Evaluation: Comparing the dierences Generally, the results of the present analysis can be evaluated as confirming the existence of gender-based dierences in apologizing. However, while some tendencies appear in both languages, the discrepancies are not always equally significant and, in a few cases, preferences associated with one of the genders in one language seem more characteristic of the other gender in the other language. As table 7 illustrates, there are more statistically significant dierences between languages than between genders, while significant gender-specific dierences are slightly more frequent in Russian but reach a higher level of significance in English. In both languages, women used more IFIDs and opted for the form with the stronger illocutionary force more often than men. The English women used the full form of the expression of regret 1.6 times as frequently as did English men, whereas Russian females employed the imperative form of the verb prostit 1.8 times as often as did Russian males.

The stronger preference for these IFID realizations by females can be interpreted as confirming Brown and Levinsons thesis that there is a systematic higher rating of FTAs by women (1987: 32). The Russian mens preference for stylistically marked IFIDs, however, stands in opposition to this pattern. Although English and Russian women agree in using more intensifiers and exclamations than men, the discrepancies are less marked in Russian. Whereas English women used 2.1 times as many intensifiers and nearly 2.5 as many exclamations as did English men, in the Russian data, these dierences amount to 1.8 in both cases. However, the much higher frequency of the pronoun menja in the female data increases the discrepancy in the use of intensification by Russian men and women. The consistently, though only slightly, higher frequency of positive politeness apology strategies in the female data seems to confirm the thesis that women put more eort into maintaining relationships than men. Since oers of repair are beneficial to the hearer, the English and Russian womens preference for direct realizations of this apology strategy can be interpreted as rendering further support to such an explanation. Furthermore, Russian women also used more appeasers and adverbials emphasizing the speakers commitment. 278 Eva Ogiermann Although English and Russian women agree in using more upgrading and fewer downgrading accounts than men, their choices of the various upgrading and downgrading account types are partly contradictory. While both English and Russian men used more excuses, the distribution of lack of intent is diametrically opposed in the two languages, with English men and Russian women displaying a particular preference for this strategy. Furthermore, English women employed expressions of embarrassment and acted innocently more often than men. The former may be associated with female speech due to their emotional character, and the latter may be interpreted as displaying a more defensive attitude than the explicit denials of responsibility favored by English males. Interestingly, the classification of the responses according to whether they entail responsibility acceptance considerably reduces the dierences Table 7. Frequencies across genders and languages: N 400 for gender groups, N 800 for language groups English J English I ENGLISH Russian J Russian I RUSSIAN IFIDs & Intensification IFID *297 *348 645 278 308 586 Intensifier ***89 ***184 ***273 **49 **87 ***136 Exclamation ***48 ***118 **166 **40 **72 **112 Positive politeness strategies Repair 172 193 365 182 199 381 Forbearance 19 28 47 25 28 53 Concern 23 32 ***55 7 13 ***20 P 214 253 467 214 240 454 Accounts Downgrading Opt out 6 6 12 5 4 9

Deny 34 22 **56 13 13 **26 Act innocently 25 36 ***61 11 7 ***18 Minimize 16 14 30 11 11 22 Excuse 42 25 67 *45 *25 70 Admit facts 32 36 **68 14 23 **37 P 155 139 ***294 99 83 ***182 Upgrading Justify 206 234 440 214 227 441 Lack of intent 37 33 70 ***26 ***59 85 Embarrassment 4 12 *16 2 2 *4 Accept 36 35 71 36 34 70 Self-criticism 2 6 8 3 4 7 P 285 320 605 281 326 607 * p < .05, ** p < .01, *** p < .001 (chi-square test) On the culture-specificity of linguistic gender di.erences 279 between genders. In both languages, the numbers of apologetic responses are nearly equal for both genders. Russian women accepted responsibility 331 times while Russian males did so 325 times. In the English data, where the dierences in the use of individual apology strategies between genders are even statistically significant, there are 297 responses accepting responsibility in the female and 292 in the male data. Clearly, an analysis of the illocutionary force of entire responses calls for a re-interpretation of the seemingly more apologetic attitude derivable from the womens stronger preference for individual apology strategies. In contrast to comparisons of total numbers of strategies identified in the data, generally portraying women as more polite, the proportions of responses accepting responsibility reveal a self-protecting tendency. Perhaps female politeness reported in previous research reflects concern for adhering to social norms and appearing polite rather than concern for the other. I am aware of the constraints on my analysis imposed by the written medium with which the data were collected. Any investigation of the speakers intention, in particular, will never be fully reliable without taking into account prosodic and kinesic features. Nevertheless, my results suggest that an analysis of written data going beyond the provision of frequencies of politeness strategies can provide valuable insights into the speakers attitude and the potential perlocution of the speech act under investigation. 8. Conclusion The analysis clearly shows that there are dierences between male and female use of apology strategies in both English and Russian, though not as many as there are between the two languages. Nonetheless, these results could be viewed as rendering support to the theory that men and women represent dierent sub-cultures and providing counterevidence to Zemskajas claims that Russian women and men do not have distinct conversational styles. On the other hand, all four groups resorted to the same range of strategiesthe dierences being mainly quantitative, so that the results could be interpreted as merely confirming the universal character of Olshtain and Cohens (1983) speech act of apologizing. The fact that British women used particularly many routine strategies

and intensifying devices rendering their behavior apologetic and polite even when accompanied by face-saving strategies seems rather striking if one considers that feminist linguists argue against the existence of genderspecific dierences (e.g., Freed 2005). 280 Eva Ogiermann It is certainly true that the position of women in British society has considerably changed over the past decades; that they have gained access to the same educational and employment opportunities as men, and that this has aected their communicative styles. However, the reason why the media continue to characterise womens and mens language as different (Freed 2005: 700) may well be that they still are. Even though recent research in feminist linguistics has amply illustrated that there is variation in womens speech, Lakos characterization of womens language continues to be accepted by diverse groups of speakers as a valid representation of their own discursive experiences (Buchholz & Hall 1995: 6). While the womens liberation movement has contributed toward more equality between genders, the changes Russia has been undergoing since the fall of the Iron Curtain have led to the development of new gender roles (described in Temkina & Rotkirch 1997). While some Russian women take on additional jobs in private companies, focus on their career, and let their parents and (remarkably less often) husbands care for their children, those whose husbands happen to have a well-paid job become housewives. Although most Western feminists would regard this development as a step back, many Russian women appreciate the time they can spend with their children and do not expect any help in the household from the familys breadwinner. These new gender roles can be expected to aect their conversational styles. The present study has examined apologies uttered by members of a predominantly female community of practice in two cultures, and it has identified a number of dierences between female and male use of apologies. These dierences can be expected to be greater in other communities of practice and across various communities of practice. Furthermore, gender roles are more distinct in some cultures than they are in others and societal changes tend to aect gender roles and womens language. Many more studies examining the relationship between language and gender as well as other aspects of social identity are necessary before the gender-specific language patterns established so far can be described as universal. While quantitative research on apologies provides valuable insights into gender-specific ways of restoring social equilibrium, qualitative studies can also shed light on the various functions apology formulae assume, and how women and men dier in employing them. Appendix English version of the questionnaire (the order was randomized for data collection) On the culture-specificity of linguistic gender di.erences 281 Your native language(s): Your age: Your gender: The subject(s) you study: Imagine yourself in the situations below and try to react as spontaneously

as possible (dont think). Please, use direct speech. Example: You are returning a book at the library and the librarian notices that you have spilled coee over it. I am terribly sorry. My little brother pushed me when I was reading in the kitchen. 1. When going on holiday your friend gave you his flat keys and asked you to feed his fish. You have not always had the time and some of the fish have died. When you return the keys your friend asks what happened. 2. Your friend had asked you to return some video tapes for her. You totally forgot and she has just received a call from the video shop, saying that the films are required by another customer and she owes a weeks fees. 3. You see a friend of yours in the crowd, run up to him and hit him on the back. Only then you realize that its not your friend, but a complete stranger. 4. When leaving a crowded shop you let go a heavy door and it hits a woman behind you. 5. You have borrowed a book from a professor. Now you are supposed to give it back to him, but you cannot remember where you put it. 6. You had a party at your flat. The next day you meet the landlady, who lives in the same house. She complains about the noise and the dirty staircase. 7. You are at a shopping centre and having an interesting conversation with your friend. You are so engaged in it that you dont realize that you are holding a CD in your hand that you were going to buy. You leave the shop and the alarm goes o. A security guy comes up to you. 8. You are just in time to catch your train and have neither a ticket nor money with you. You have just taken a seat when the ticket inspector enters the compartment. She asks you for your ticket. Distracter 1. You have lent a book to a friend of yours and she returns it in a bad condition. 282 Eva Ogiermann Distracter 2. You got ill and cannot attend an important lecture. You ring up a fellow student to ask if you can copy his notes. Notes 1. A provisional version of this paper was pre-published as: A Contrastive Analysis of Gender-Based Dierences in English and Russian Apologies by LAUD (eds.), Series A: General & Theoretical Papers, Paper No. 669. Duisburg-Essen University and presented at the 31st International LAUD Symposium. The preparation of the final version involved a new categorization of the data. I would like to thank the conference participants who attended my lecture for the interesting discussion and the two anonymous reviewers for their helpful comments. The remaining errors are all mine.

2. For methodological reasons, addressee gender was not included in the analysis. An examination of DCT data collected in languages in which gender is encoded morphologically, such as Russian and Polish, has shown that addressee gender can be easily misinterpreted. Moreover, the situations used in the questionnaire are all dierent and include several factors other than gender, making it dicult to extract it as the variable responsible for strategy choice. In retrospective interviews conducted with some of the subjects, gender was seldom named as the variable leading to the preference of one strategy over another. 3. For classification criteria, examples and discussion of problematic cases see Ogiermann (2007). References Aijmer, Karin. 1996. Conversational Routines in English: Convention and Creativity. London: Longman. Azgichina, Nadezda. 2000. Na puti k obreteniju sily (On the way to getting strength). Accessed at: http://www.owl.ru/win/info/we_my/2000_sp/05.htm (1/11/07) Bataineh, Ruba Fahmi, and Rula Fahmi Bataineh. 2005. American university students apology strategies: An intercultural analysis of the eect of gender. Journal of Intercultural Communication 9. Accessed at: http://www.immi.se/intercultural/nr9/bataineh.htm (1/11/07) Bataineh, Ruba Fahmi, and Rula Fahmi Bataineh. 2006. Apology strategies of Jordanian EFL university students. Journal of Pragmatics 38: 19011927. Bean, Judith Mattson, and Barbara Johnstone. 1994. Workplace reasons for saying youre sorry: Discourse task management and apology in telephone interviews. Discourse Processes 17: 5981. Beebe, Leslie M., and Martha Clark Cummings. 1996. Natural speech data versus written questionnaire data: How data collection method aects speech act performance. In Susan M. Gass and Joyce Neu (eds.), Speech Acts Across Cultures. Berlin: Mouton de Gruyter. 6586. Bergvall, Victoria L. 1999. Toward a comprehensive theory of language and gender. Language and Society 28: 273293. Blum-Kulka, Shoshana, Juliane House, and Gabriele Kasper (eds.). 1989. Crosscultural Pragmatics: Requests and Apologies. Norwood, N. J.: Ablex. On the culture-specificity of linguistic gender di.erences 283

Brown, Penelope, and Stephen C. Levinson. 1987. Politeness. Some Universals in Language Usage. London: CUP. Bucholtz, Mary. 1996. Black feminist theory and African American womens linguistic practice. In Victoria L. Bergvall, Janet M. Bing, and Alice F. Freed (eds.), Rethinking Language and Gender Research: Theory and Practice. London: Longman. 267290. Bucholtz, Mary. 1999. Why be normal?: Language and identity practices in a community of nerd girls. Language in Society 28: 203223. Buchholz, Mary, and Kira Hall. 1995. Introduction: 20 years after Language and Womans Place. In Kira Hall and Mary Buchholz (eds.), Gender Articulated: Language and the Socially Constructed Self. New York/London: Routledge. 122. Christie, Chris. 2005. Politeness and the linguistic construction of gender in parliament: An analysis of transgressions and apology behavior. Accessed at: http://extra.shu.ac.uk/wpw/ politeness/christie.htm (1.11.2007) Coates, Jennifer. 1994. The language of the professions: Discourse and career. In Julia Evetts (ed.), Women and Career: Themes and Issues in Advanced Industrial Societies. London: Longman. 7286. Coates, Jennifer. 1996. Women Talk. Oxford: Blackwell. Comrie, Bernard, and Gerald Stone. 1978. The Russian Language since the Revolution. Oxford: Clarendon Press. Cordella, Marisa. 1990. Apologizing in Chilean Spanish and Australian English: A crosscultural perspective. Australian Review of Applied Linguistics 7: 6692. Deutschmann, Mats. 2003. Apologising in British English. Umea : Umea Universitet. Doleschal, Ursula, and Sonja Schmid. 2003. Doing gender in Russian: Structure and perspective. In Marlis Hellinger and Hadumod Bumann (eds.), Gender Across Languages II. Amsterdam/Philadelphia: John Benjamins. 253282. Eckert, Penelope, and Sally McConnell-Ginet. 1998. Communities of Practice: Where Language, Gender, and Power All Live. In Jennifer Coates (ed.), Language and Gender. Oxford: Blackwell. 484494. Eelen, Gino. 2001. A Critique of Politeness Theories. Manchester: St. Jerome Publishing. Fillmore, Charles J. 1971. Verbs of judging: An exercise in semantic description. In Charles J. Fillmore and D. Terence Langendoen (eds.), Studies in Linguistic Semantics. New York: Holt, Rinehart & Winston. 273289. Fishman, Pamela. 1983. Interaction: The work women do. In Barrie Thorne, Cheris Kramarae,

and Nancy Henley (eds.), Language, Gender and Society. Boston, Mass.: Heinle & Heinle. 89102. Formanovskaja, Natalja. 2002. Russkij Recevoj E tiket. Normativnyj Socjokulturnyj Kontekst. Moskva. Russkij Jazyk. Freed, Alice F. 2005. Epilogue: Reflections on language and gender research. In Janet Holmes and Miriam Meyerho (eds.), The Handbook of Language and Gender. Oxford: Blackwell. 699721. Goodwin, Marjorie Harness. 2005. The relevance of ethnicity, class, and gender in childrens peer negotiations. In Janet Holmes and Miriam Meyerho (eds.), Handbook of Language and Gender. London: Blackwell. 229251. Holmes, Janet. 1990. Apologies in New Zealand English. Language in Society 19: 155199. Holmes, Janet. 1995. Women, Men and Politeness. New York: Longman. Holmes, Janet. 2005. Politeness and postmodernisman appropriate approach to the analysis of language and gender? Journal of Sociolinguistics 9: 108117. Kasper, Gabriele. 2000. Data collection in pragmatics research. In Helen SpencerOatey (ed.), Culturally Speaking: Managing Rapport through Talk across Cultures. London: Continuum. 316341. 284 Eva Ogiermann Koniuszaniec, Gabriela, and Hanka Baszkowska. 2003. Language and Gender in Polish. In Marlis Hellinger and Hadumod Bumann (eds.), Gender Across Languages III. Amsterdam/Philadelphia: John Benjamins. 259285. Lako, Robin. 2004. Language and womans place. In Mary Buchholtz (ed.), Language and Womans Place. Text and Commentaries. Revised and Expended Edition. Oxford: OUP. 39102. Liljestrom, Marianne. 1995. Emanciperade till Underordning. Det Sovjetiska Ko nssystements Uppkomst och Diskursiva Reproduktion. Turku: A bo Akademi University Press. Maltz, Daniel N., and Ruth A. Borker. 1982. A cultural approach to male-female miscommunication. In John Gumperz (ed.), Language and Social Identity. Cambridge: CUP. 195216. Marquez Reiter, Rosina. 2000. Linguistic Politeness in Britain in Uruguay: A Contrastive Study of Requests and Apologies. Amsterdam: John Benjamins. McElhinny, Bonnie, S. 1998. I dont smile much anymore: Aect, gender, and the discourse of Pittsburgh police ocers. In Jennifer Coates (ed.), Language and Gender. A Reader. Oxford: Blackwell. 309327.

Mendoza-Denton, Norma. 2004. The anguish of normative gender: Sociolinguistic studies among U.S. Latinas. In Mary Buchholtz (ed.), Language and Womans Place. Text and Commentaries. Revised and Expended Edition. Oxford: Oxford University Press. 260 268. Meyerho, Miriam. 1999. Sorry in the Pacific: Defining communities, defining practices. Language in Society 28: 225238. Mills, Margaret H. (ed.). 1999. Slavic Gender Linguistics. Philadelphia: John Benjamins. Mills, Sara. 2003. Gender and Politeness. Cambridge: Cambridge University Press. Ogiermann, Eva. 2002. Gender-based dierences in English apology realisations. Paper presented at the Conference on the Pragmatics of Interlanguage English, Mu nster. To appear in Ronald Geluykens and Bettina Kraft (eds.), Cross-Cultural Pragmatics and Interlanguage English. Muenchen: Lincom Europa. Ogiermann, Eva. 2004. Review of Mats Deutschmanns Apologising in British English. Umea 2003. Journal of Sociolinguistics 8: 302306. Ogiermann, Eva. 2006. Cultural variability within Brown and Levinsons politeness theory. English, Polish and Russian apologies. In Cristina Mouron Figueroa and Teresa Iciar Moralejo-Garate (eds.), Studies in Contrastive Linguistics. Santiago de Compostela: Universidade de Santiago de Compostela Publicacio ns. 707718. Ogiermann, Eva. 2007. Politeness in Britain, Poland and Russia. A Contrastive Analysis of Apologies. Unpublished doctoral dissertation, University of Oldenburg, Germany. Olshtain, Elite and Andrew D. Cohen. 1983. Apology: A speech act set. In Nessa Wolfson and Elliot Judd (eds.), Sociolinguistics and Language Acquisition. Rowley, Mass.: Newbury House Publishers, Inc. 1835. Rathmayr, Renate. 1996. Pragmatik der Entschuldigungen. Vergleichende Untersuchung am Beispiel der Russischen Sprache und Kultur. Ko ln: Bohlau Verlag. Shardakova, Maria. 2005. Intercultural pragmatics in the speech of American L2 learners of Russian: Apologies oered by Americans in Russian. Intercultural Pragmatics 2: 423 451. Stapleton, Laura Elizabeth. 2004. Variation in the performance of speech acts in Peninsular Spanish: apologies and requests. Dissertation: Louisiana State University and Agricultural and Mechanical College. Accessed at: http://etd.lsu.edu/docs/available/etd-06032004191432/unrestricted/Stapleton_dis.pdf

Suszczynska, Magorzata. 1999. Apologizing in English, Polish and Hungarian: dierent languages, dierent strategies. Journal of Pragmatics 31: 10531065. On the culture-specificity of linguistic gender di.erences 285 Suszczynska, Magorzata. 2005. Apology Routine Formulae in Hungarian. Acta Linguistica Hungarica 52: 77116. Tafel, Karin. 1997. Die Frau im Spiegel der russischen Sprache. Wiesbaden: Harrassowitz Verlag. Tannen, Deborah. 1994. Talking from 9 to 5: Women and Men at Work. Language, Sex and Power. London: Virago Press. Temkina, Anna, and Anna Rotkirch. 1997. The fractured working motherand other new gender contracts in contemporary Russia. Accessed at: http://www.valt.helsinki.fi/sta/ rotkirch/gendcontract.htm (1.11.2007) Thorne, Barry, and Nancy Henley (eds.). 1975. Language and Sex: Di.erence and Dominance. Rowley, Mass., Newbury House Publishers. Watts, Richard. 2003. Politeness. Cambridge: Cambridge University Press. Zemskaja, Elena, Margarita Kitajgorodskaja, and Nina Rozanova. 1993. Osobennosti muzskoj i zenskoj reci. In: Elena Zemskaja and Dmitrij S melev (eds.), Russkij Jazyk v ego Funkcionirovanii. Kommunikativno-Pragmaticeskij Aspekt. Moskva: Nauka. 90 135.